Tokyo AFM

Prior to joining Tokyo Auto Fire and Marine (Tokyo AFM), a publicly traded Japanese casualty

insurance company, in June 2001, Nobu Matsumoto had held various management positions in the

insurance industry for 20 years, in both Japan and overseas. He was appointed chief executive officer

of Tokyo AFM after two financial service companies from the U.S. and Europe each acquired a 20%

interest in Tokyo AFM. The two new investors intended to rapidly expand the operations of the

company overseas.

Industry Background

Casualty insurance companies had two principal sources of revenue. The first source was

insurance premiums, which were payments that clients (hereafter policyholders)individuals or

businessesmade to insurance companies to provide protection against losses resulting from

adverse events such as fire or natural catastrophes. Typically, these premiums were paid up front in

cash for protection covering one-year to five-year periods. The main costs associated with insurance

contracts arose from the actual payment of losses that policyholders subsequently incurred and that

were covered by the contracts.1 Loss-related expenses were mainly divided between direct claim

payments and indirect expenses related to processing claims. Typically, insurers also incurred costs

to acquire customers and set up policies. The main categories of acquisition costs consisted of

commissions paid to agents and salespeople, administrative policy issuance costs, advertising

expenditures, and agent recruitment and training.

The second source of revenue for insurance companies was investment income. Insurers invested

their float in various financial instruments. In simple terms, the float was the amount of cash

collected from policyholders and not yet paid out for claims or other expenses. Since casualty,

particularly catastrophic, losses could occur at any time, one of the challenges for insurers was to

maintain an adequate level of liquidity in their portfolio to be ready to pay claims as they arose.

Brief Company History

Tokyo AFM was established in Tokyo in 1928 as Nippon Insurance Co., Ltd., which specialized in

property fire-damage insurance. Tokyo AFM gradually widened the range of its products over time

to become a more comprehensive property-casualty insurance group. The company was listed on the

Tokyo Stock Exchange in 1963. Over the years, the companys profits had grown at a slow but steady

pace until the casualty insurance industry was deregulated in the late 1990s. Soon after, the financial

performance of Tokyo AFM deteriorated.

Despite Tokyo AFMs desire to remain an independent insurer, the industrys deregulation

proved challenging. In early 2001, The American Banking Group acquired a 23.04% stake in Tokyo

AFM, and the German reinsurance group Bayern Re acquired 20.54% of the companys shares.

Soon after his appointment as CEO, Matsumoto became concerned that certain financial

accounting policies of the company did not reflect the economic reality of the underlying

transactions, particularly those related to revenue recognition, contract acquisition costs, reserves for

contingent future losses, and investments in marketable securities. He asked that you comment on

the companys current accounting practices and suggest any changes you might recommend, along

with your reasons. Be sure to identify the alternatives you rejected and your reasons for rejecting

them.2 Do not dismiss an alternative or reach a decision on the grounds of immateriality. If you

make any assumptions, please state them.

3. Broadly speaking, two major types of insured events could give rise to losses covered by

insurance contracts:

Events that actuarial analysis could assess and predict with a high level of accuracy across a

large number of contracts (for example, events covered by automobile insurance).

Catastrophes, which were generally adverse natural events such as earthquakes and

hurricanes, but which could also be human-induced events such as terrorist attacks.

Catastrophes were considered low probabilityhigh consequence events. They were

uncertain and very difficult to predict in terms of timing and extent of damage.

For the coming year, the company had estimated that expected losses across all its automobile

insurance contracts would amount to 70% of premiums. (Exhibit 1 shows historical data on losses

incurred by Tokyo AFM in automobile insurance and catastrophe insurance.)3 With respect to

catastrophes, Matsumoto had concluded that the average expected losses would be 12% of

premiums, but that any scenario between 4% and 20% was equally likely.4

Question What accounting treatment would you choose for expected losses (a) associated

with automobile contracts and (b) associated with catastrophes? From a shareholders perspective,

what concerns do you think could arise with respect to the accounting treatment of expected losses?

4. Tokyo AFM accounted for its portfolio of investments in marketable securities at historical cost.

The company invested primarily in long-term, liquid financial instruments (typically five-year

bonds). Although he expected that a large portion of its portfolio would be held until the bonds

matured, the CEO wanted to set aside the remainder as ready to be sold at any time. Indeed, adverse

events triggering losses could occur any day, and the company needed to maintain a certain level of

liquidity to meet immediate cash needs.

Question How would you account for the companys marketable securities? (See Exhibit 2 for

price data on a 100,000-bond representative of Tokyo AFMs investments.)

Exhibit 1 Tokyo AFM's Historical Loss Ratioa for Automobile Insurance and Catastrophe-related Damages 80.0% 70.0% 60.0% 50.0% 40.0% 30.0% 20.0% 10.0% 0.0% + Automobile Catastrophes 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Exhibit 2 Holdings Market Price for Five-Year AAA Prime Credit Bond, Representative of Tokyo AFM's 108 106 104 Dec. 31, 2000: 100 102 Dec. 31, 1996: 105 100 98 Dec. 31, 1998: 101 96 Dec. 31, 1997:98 94 92 Dec. 31, 1999:95 90 88 Jul-96 Jan-97 Apr-97 Jul-97 Oct-97 Jan-98 Apr-98 Jul-98 Jan-99 Apr-00 Oct-00 Jan-96 Apr-96 Oct-96 Oct-98 Apr-99 Oct-99 Jan-00 Jul-00